UC Berkeley

Smaller, lighter, more powerful batteries will help drive the next generation of mobile technologies with energy from renewable sources. Battery chemistries incorporating a lithium metal anode offer high theoretical specific energies and energy densities, but no current electrolyte has yet been shown to be adequate, despite considerable effort.

Lithium metal electrodes suffer from irreversibility during battery cycling which diminishes their utility. First, lithium metal is highly reactive and tends to consume electrolyte with irreversible chemical reactions at electrode/electrolyte interfaces. Second, lithium metal tends to plate and strip in a nonplanar fashion, resulting in lithium protrusions that may span the electrolyte and short-circuit the cell. This is the so-called “dendrite problem”, named for the branched, filamentous lithium structures first observed on lithium metal after cycling in liquid electrolytes. Theoretical models suggest that planar deposition of lithium can be encouraged by using a rigid electrolyte, wherein compressive forces on the electrode discourage the growth of protrusions. However, the non-negligible displacement of the electrode/electrolyte interface as the lithium electrode is charged and discharged places additional constraints on the electrolyte: if the electrolyte cannot maintain good electrochemical contact with the electrode, it is unsuitable. Thus, it is desirable for an electrolyte to have some solid-like properties (e.g. resistance to lithium protrusion growth) and some liquid-like properties (e.g. able to accommodate of volume change). The viability of next generation lithium and beyond-lithium battery technologies hinges on the development of electrolytes with improved performance.

A suitable electrolyte must promote reversible lithium metal electrodeposition and stripping, and also conduct ions at a sufficiently high rate. Nanostructured electrolytes have shown promise for this purpose: such materials contain conducting domains for ion transport and rigid domains to promote stable lithium deposition. This dissertation focuses on model nanostructured block copolymer electrolytes comprising polystyrene-block-poly(ethylene oxide) (PS-b-PEO, or SEO) copolymers mixed with a lithium salt. By changing total chain length and the ratio of PS to PEO, one can obtain electrolytes with a variety of self-assembled morphologies and a range of mechanical properties. The salt is preferentially located in the PEO-rich lamellae. We focus on those electrolytes that self-assemble into lamellar morphologies, with alternating layers of ion-conducting PEO and mechanically reinforcing PS. Such composite materials have ion transport properties that are not easily predicted from the ion transport properties of the conducing domain only.

The design of new solid electrolyte materials is currently limited by incomplete fundamental understanding of the relationship between the initial properties of the electrode and electrolyte (e.g. mechanical, ion transport), and the resulting morphology of electrodeposited lithium after cycling. To overcome this challenge, this dissertation includes studies of both fundamental ion transport and lithium morphology in the same electrolyte. The morphology of electrodeposited lithium is affected by many factors, and lithium electrodeposition through different electrolytes has largely been studied qualitatively. We study lithium deposition by analyzing three-dimensional (3D) position space images of lithium deposition using synchrotron hard X-ray tomography. 3D renderings provide an opportunity to quantitatively correlate various factors such as experimental current density, interelectrode distance, and electrode shape and understand how lithium protrusion morphology evolves over time under conditions of plating and stripping. Ion transport through an electrolyte and lithium electrodeposition are inextricable. We also study lithium ion transport through a series of lamellar SEO electrolytes.

Chapter 1 provides an introduction to topics of interest to this dissertation, including lithium metal batteries and associated challenges, electrolytes for lithium batteries, nonplanar electrodeposition of lithium metal, and information about the imaging technique X-ray tomography. Current density, properly normalized, is perhaps the key variable in the electrodeposition of lithium. Chapter 2 is the first attempt at quantifying the effect of current density on the geometry and density of dendrites and other protrusions during electrodeposition through a solid polymer electrolyte. Lithium protrusions in cells with SEO electrolytes have been observe to nucleate on impurity particles in the existing lithium metal electrode. In Chapter 3, an “electrochemical filtering” treatment was performed on lithium symmetric cells in order to reduce the concentration of impurity particles near the electrode-electrolyte interface, and cells were cycled to determine the effects of the treatment on lifetime. In Chapter 4 we demonstrate that globules are preferentially stripped compared to a planar electrode in our system, which incorporates a nanostructured block copolymer electrolyte. First, we show and quantify the healing of lithium metal anodes during stripping; second, we also view protrusions from a moving reference frame attached to the electrode/electrolyte interface. Chapter 5 aims to quantitatively analyze the plating and stripping of a lithium protrusion in a lithium metal symmetric cell over a series of time steps based on X-ray tomography imaging. Current density at the positive electrode, current density at the negative electrode, and interelectrode distance were quantified and mapped as a function of position. Correlation functions are calculated to reveal the relationships between these three microscopic measured parameters. These results shed light on the interactions between parallel electrodes separated by a battery-relevant distance. Chapter 6 deals with ion transport in a series of twelve SEO block copolymer electrolytes, all with nominally lamellar morphology. These electrolytes include those through which lithium electrodeposition was studied in Chapters 2, 3, 4, and 5. Ionic conductivity, salt diffusion coefficient, current fraction, interfacial impedance, and limiting current density were measured in these electrolytes and compared to those of homopolymer PEO. Scanning electron microscopy was used to characterize the self-assembled morphology of these electrolytes at different salt concentrations. We compare measured limiting current density in the twelve nanostructured electrolytes with predictions based on the effective-medium theory of Sax and Ottino and Newman’s concentrated solution theory.

To make progress towards commercially viable Li metal batteries, a deep, fundamental understanding of lithium deposition that simultaneously considers chemical, mechanical, and electrochemical factors will be crucial. The objective of this work is to increase our understanding of these interactions in order to inform models and materials design to enable next-generation electrochemical energy storage.